Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Chapter 22
Preliminary Results on Biomimetic Methods Based on Soluble Ammonium Phosphate Precursors for the Consolidation of Archaeological Wall Paintings Magdalena Balonis-Sant,*,1 Xiao Ma,1 and Ioanna Kakoulli1,2 1Department
of Materials Science and Engineering, University of California Los Angeles, 410 Westwood Plaza, 3111 Engineering V, BOX 159510, Los Angeles, California 90005-1595 2UCLA/Getty Conservation Program and Cotsen Institute of Archaeology, University of California Los Angeles, 308 Charles E. Young Drive North, A210 Fowler Building, Box 951510, Los Angeles, California 90095-1510 *E-mail:
[email protected] This research develops hydroxyapatite (HAP)-based, inorganic mineral systems for the consolidation of powdery wall paintings of archaeological significance. The scientific approach exploits biomimetic (biologically inspired design) principles to induce in situ the formation of protective HAP crystals by triggering reactions between the calcium carbonate-rich layers in wall paintings and ammonium phosphate precursors. The high solubility and absence of toxicity of ammonium phosphate precursors and the stability of the hydroxyapatite reaction product at varying pH, renders this treatment extremely promising for consolidation and protection of weathered wall paintings. Experimental trials were carried out on wall painting test blocks applying cellulose compresses of 1M and 2M solutions of diammonium hydrogen phosphate for 3 to 6 hours contact time. The consolidating effect was evaluated through microstructurally and compositionally-sensitive analytics including VPSEM-EDS, XRF, water sorption test and scotch tape test. Preliminary results indicated the formation of a porous hydroxyapatite network at the surface and subsurface of the test blocks with improved cohesion, pH-resistivity and © 2013 American Chemical Society In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
reduced water absorption. These data show the potential of this treatment for the consolidation of powdery wall paintings and their protection from weathering and deterioration induced by natural aging and environmental action-linked effects.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Introduction Wall paintings provide a testimony of artistic, cultural, and intellectual developments and are often of great archaeological, historical and cultural significance. Physically and chemically, wall paintings are highly complex, heterogeneous composites consisting of a layered structure of at least three phases: binder, aggregate and pigment (1). Owing to their nature and context, wall paintings are faced with imminent threats of deterioration both natural and manmade with broad impact causing chemical, physical and mechanical damage to the constituent materials. These include but are not limited to: moisture, microbiological growth, pollution, seismic activity, vandalism and others. The most common manifestations of weathering on wall paintings include staining of microbiological or chemical origin and color alteration, pitting and severe powdering leading to loss of cohesion within the different layers at the microscopic scale. Extensive studies have been carried out using a variety of consolidation treatments to improve the condition of surfaces (2–7). These studies critically indicate that choosing a proper consolidant for porous materials is challenging. This is specifically related to being able to ensure sufficient penetration beneath the surface, while also providing mechanical strength and abrasion resistance (2, 7, 8). The definition of a consolidant provided by Warren can be used as a point of reference (9): A consolidant acts at the near-molecular level by fixing or inhibiting the capacity for movement between very small particles, thereby altering the characteristics of the material in terms of its behavior, particularly in the presence of water. It tends to make the material stronger in compression and tension, and may affect inherent characteristics, such as heat and sound transmission and rigidity. (121) Further, a consolidant should not severely alter the original “nature of the porosity” (10), and not be extremely hydrophobic, which could lead to further damage. Ideally, the consolidant should also be photochemically stable to prevent optical changes of the surface of the paint that could affect significantly the saturation of color, gloss and texture (8). Also, one of the key factors in the consolidation of wall paintings in particular is retreatability, enabling future treatments, if necessary (11). Polymers, mainly acrylic, vinyl and silicone based, have been widely used as consolidants for cultural heritage objects (12–17). However, contrary to expectation, in many cases polymers used for the consolidation of wall paintings have often induced further degradation of the artwork by changing greatly the 420 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
physicochemical properties of the plasters and paint layers (18). An alternative method to organic polymers is the in situ formation of more compatible inorganic mineral phases re-establishing the loss of cohesion in disintegrated plaster layers (3, 7, 19–26). An example is the application of Ca(OH)2-based consolidants that has been practiced extensively (3, 7, 27, 28). The consolidation mechanism is based on the transformation of Ca(OH)2 into calcium carbonate (CaCO3) through a carbonation process induced by the CO2 in the atmosphere through a chemical reaction similar to the setting of the original plaster. However, the low solubility of calcium hydroxide in water (1.73 g/L at 20°C), the low stability of the system and high surface tension of water (as the medium) cause aggregation and phase separation resulting in inefficient penetration, poor consolidation strength and whitening of the surface. Several strategies including “sucrose methods” to increase Ca(OH)2 solubility (7, 29) and the synthesis of Ca(OH)2 nanoparticle-alcohol dispersions to prevent agglomeration and improve penetration have been reported (21). Yet, in spite of improved working and performance properties, the sensitivity of CaCO3 to a drop in the pH (caused by acid rain and other acid local environments for example) or a phase transformation into gypsum (CaSO4.2H2O) in the presence of sulfate ions (known as ‘sulfation’) poses limitations in the use of Ca(OH)2-based consolidants. Other inorganic mineralization treatments based on barium or oxalate salts have been questioned due to concerns related to the toxicity of barium (not all nations permit the use of barium in conservation) and the limited penetration of oxalates into a porous medium (3, 30, 31). Recent studies (31–35) have demonstrated a considerable potential for developing much improved consolidation methods by bio-mimicking the growth of hydroxyapatite (Ca10(PO4)6(OH)2, HAP), the main mineralogical component of teeth and bones on limestone matrices (31, 32, 34, 36). In this research, we explore similar principles to induce the in situ formation of protective HAP crystals by triggering reactions between the Ca in the calcium carbonate (CaCO3)-rich plaster layers of wall paintings and ammonium phosphate precursors. Both monoammonium (MAP) and diammonium hydrogen phosphate (DAP) precursors were tested though here only the use of the latter is discussed. Other soluble phosphate-based salts such as potassium/sodium phosphates could also serve as a source of phosphorous. However, in conservation applications it is not advisable to use Na+ and K+ cations as upon penetration they may combine with anions present in the system (or deposited from the atmosphere) to form Na and K salts that could be detrimental to the preservation of the wall paintings (37, 38). The theoretical chemical pathway of the HAP formation using DAP as the precursor is presented below (Reaction 1) (35) and a schematic diagram of the anticipated consolidation effect in Figure 1.
In reality, the HAP formed through this reaction may be non-stoichiometric; Ca deficient or HAP containing substituted carbonate species may form (34, 35, 39, 40). The HAP precipitation is expected to be preceded by the formation 421 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
of intermediate metastable phases, such as amorphous calcium phosphate, monocalcium phosphate monohydrate, dicalcium phosphate dehydrate and/or octacalcium phosphate. These intermediate phases whose formation depends on the reaction conditions (e.g., degree of supersaturation, temperature, pH, presence of foreign ions, etc.), are expected to transform into HAP or its ion-substituted analogues (34, 36, 39, 41, 42). The resulting hydroxyapatite network is expected to improve cohesion between loose particles at the immediate subsurface of wall painting. Since hydroxyapatite is reported to be stable in a wide pH range (between ~pH 4 and 14) (43), increased resistivity towards an acid environment is anticipated (44).
Figure 1. Schematic representation of the in situ consolidation of a wall painting plaster. HAP (black dotted surface) forms on the top of the surface treated, preferentially in areas occupied by microcrystalline CaCO3 (plaster binder). The large grey/whitish particles represent aggregates in the plaster layer. Not in scale.
Experimental Methods Materials for the Preparation of Wall Painting Test Blocks Wall painting plaster test blocks were used to test the hydroxyapatite (HAP) consolidation effect. The test blocks were prepared by mixing 1) commercial calcium hydroxide (lime putty) with marble dust (0.2-0.6 mm) and deionized water at 1:4:1 volumetric ratio and 2) commercial lime putty with sand (0.2-0.6 mm) at the same ratio as above. The mixtures at a consistency of a dense paste (mortar) were subsequently placed in custom-made stainless-steel molds (5cm x 5cm x 2cm) and left to dry at 25° C and ~60% RH for one week. Consequently the samples were de-molded and left to set fully for 6 months in order to ensure full conversion of the Ca(OH)2 binder to CaCO3. The lime putty, marble dust and sand were purchased from Kremer Pigments Inc. 422 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Application and in Situ Formation of HAP For the application of the reagent, cellulose compresses (poultices) were prepared of 1 and 2 M DAP. Each compress was approximately 1 cm thick. A permeable very thin tissue of compact weave was placed at the interface between the wall painting test block and the cellulose poultice to prevent cellulose fibers from sticking on the surface of blocks (Figure 2). Three and six hours of contact time were tested. Cellophane foil was used to cover the cellulose compress in order to prevent the solution from evaporating too fast. During the reaction, a faint smell of ammonia was detected. After completion of the treatment, the cellulose poultice was removed and the test blocks were left to dry in air. The DAP solutions were made using deionized water and analytical reagent grade diammonium hydrogen phosphate purchased from Fisher Scientific. Each cellulose compress was prepared by mixing Arbocel cellulose fibers: BC 200 and BWW 40 at a volumetric ratio of 0.4 (BC 200) : 0.6 (BWW 40) : 0.9 (DAP solution). Tengucho tissue purchased from Hiromi Paper, Inc. was used as the intermediate supporting layer.
Figure 2. Example of an application of the consolidation treatment on the wall painting test blocks (a: schematic, b: actual).
Determination of Carbonation Level of Simulated Wall Painting Substrate Test Blocks The calcium carbonate/calcium hydroxide content was determined using thermal analysis (TG/DTG) performed in a nitrogen environment using a Perkin Elmer STA 6000 thermal analyzer. Temperatures were scanned in the range between 35-980°C and heating rate was set for 10°C/min. 423 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Morphological and Physicochemical Characterization of HAP and Assessment of Treatment Efficiency The effectiveness of the treatment in providing consolidation action and the formation of HAP network was evaluated using scanning electron microscopy (SEM) at variable pressure (VP) coupled with energy dispersive x-ray spectroscopy (EDS); X-ray fluorescence spectroscopy (XRF); water sorption test and scotch tape test.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
SEM-EDS SEM analysis was performed using the FEI Nova™NanoSEM 230. To avoid the necessity of carbon or metal sputtering, a low vacuum mode (variable pressure) was applied during the analysis. Morphological and topographic characteristics of the surface before and after treatment were recorded using the secondary electron detector (SE). The depth of penetration and spatial localization of the newly phosphate-containing forms was assessed using backscattered electron detector (BSE) and energy dispersive x-ray spectroscopy (EDS) on polished sections. Polished sections were prepared by embedding samples in epoxy resin (Buehler EpoxiCure®) mixed with EpoxiCure® Epoxy Hardener and placed under vacuum using the Buehler Cast n’ Vac 1000. Once the resin had set, the samples were cut using Labcut 1010 low speed diamond saw and ground using Buehler silicon carbide grinding papers from 240 to 1200 grit. Samples were subsequently polished on a Leco® GP-25 polishing turntable using water-based diamond suspensions of 6 μm and 1 μm followed by colloidal silica of 0.2 μm spread on Buehler® MasterTex polishing cloths.
XRF XRF data provided complementary semi-quantitative results on the phosphorous content using the Thermo Scientific Niton® XL3t Series GOLDD™ technology handheld XRF analyzer with a silver anode and silicon drift detector. As all the consolidation trials were performed on the test blocks positioned vertically, it was anticipated that measured P content would be higher for the slices taken from the bottom of the sample due to the gravity, while slightly lower for the slices taken from the top of the wall painting test blocks. To have the most representative and consistent information sections from the center of the samples were chosen for the preliminary XRF analysis. More detailed XRF information on the gravity effect will be conducted in the scope of future analysis.
424 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Water Sorption Test
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
The water sorption test was conducted following a modified version of the ASTM standard C1585. For this test both untreated and treated test blocks were kept in the atmospheric air at 20 ± 3°C at around 60% RH (relative humidity) for the period of 7 days prior to testing. All sides of each test block (except the surface receiving the treatment) were sealed tightly with duct tape. The surfaces of untreated and treated samples were placed in contact with the water (immersed between 1-3 mm) and the mass of each sample was monitored to the nearest 0.01 g at the intervals specified in the standard. The absorption, I was calculated according to the equation 1.
where: I = the absorption mt = the change in specimen mass in grams, at the time t a = the exposed area of the specimen, in mm2 d = the density of the water in g/mm3
Scotch Tape Test The scotch tape test (adhesion test) was performed based on a modified version of the ASTM standard D3359. Strips of scotch tape (4 cm long) were pressed onto the surface of each sample using enough fingertip pressure to remove all visible air bubbles and make good contact between the tape (adhesive side) and the surface. The tape was then peeled off in a single smooth motion. Systematic mass analysis of the tape before and after adhesion with the mortar surface provided an indication of the consolidation treatment performance. The results were evaluated visually and quantitatively based on the amount of material separated from the surface before and after the consolidation treatment.
Results and Discussion Simulation of Wall Painting Substrate Test Blocks Thermal analysis (TG/DTG) data revealed that after 6 months in air and at room temperature (~25°C) the surface of the test blocks was completely carbonated whereas the core was carbonated at ~96% (±1%). Example of the TG/DTG evaluation is presented in Figure 3 (a,b) and Table I (a), I (b). Mass losses were determined from the TG curve by the points of deviation for the tangent lines to the curve observed for the weight loss versus temperature. Peak recorded between around 390°C and 550°C corresponds to the dehydration of Ca(OH)2 to CaO and H2O, while weight loss recorded between 600°C and 950°C comes from the decomposition of CaCO3 to CaO and CO2. 425 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 3. TG/DTG characteristics of test block prepared from marble dust and lime putty after 6 months of carbonation in the air: sample taken from the core; (a) TG (thermogravimetric analysis)- mass loss versus temperature; (b) DTG derivative of mass loss.
426 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Table I (a). Calculation of Ca(OH)2 content from the TG/DTG curves (Figure 3) Temperature (°C)
Sample Mass (mg)
% Mass
390
75.20
99.62
500
74.23
98.35
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Ca(OH)2 Content (%)
5.2 (±1%)
Table I (b). Calculation of CaCO3 content from the TG/DTG curves (Figure 3) Temperature (°C)
Sample Mass (mg)
% Mass
600
74.21
98.32
950
42.39
56.16
CaCO3 Content (%)
95.8 (±1%)
Morphological and Physicochemical Characterization before and after Treatment SEM-EDS The morphology and topography of the surface of the wall painting test blocks before and after the consolidation treatments was characterized using scanning electron microscopy. Secondary electron micrographs of the control (untreated) samples show large grain-aggregates of marble dust or quartz within a microcrystalline CaCO3 matrix formed through the carbonation of quenched lime (slaked lime, Ca(OH)2) used as the binder (Figure 4 (a,b)). The treated samples show the formation of calcium phosphate phases (Figures 5 (a,b) and 6 (a,b)) as a result of the in situ reaction between the DAP solution and the Ca mainly from the microcrystalline CaCO3 binder, though the Ca from the larger marble dust particles has also contributed in the formation of calcium phosphate phases. Elemental analysis using EDS of the Ca-P phases on unmounted samples, though provided semi-quantitative indication on the element wt% content, it enabled some rough evaluations to be made. The P content measured by EDS in phosphorous-rich regions (at spatial resolution of ~1 μm) ranged between ~ 3-6 wt% for 1 M treatments (Table II(a), II(b)) and ~ 2-5 wt% for 2 M treatments (Tables III(a), III(b) and IV(b)). DAP consolidation treatments at 1M concentrations seem to result in precipitation of a much more homogenous network of calcium phosphate crystals (Figure 5 (a,b)) compared to the 2 M DAP treatments (Figure 6 (a,b)). In the test blocks prepared with marble dust 427 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
in particular (but also those prepared with quartz aggregate), large crystals of a different Ca-P phase were also present (Figure 7 (a,b)). The P content of this phase ranged between ~9 and 17.5 wt% (Table IV(a), IV(b)). Traces of magnesium (Mg) were also identified. Though it has not been directly confirmed, it is thought that the newly formed calcium phosphate phase using 1 M DAP is in fact hydroxyapatite. Taking into consideration the experimental conditions (25°C and pH~8 which corresponds to 1M DAP), HAP seems to be thermodynamically the most favored phase to form due to its low solubility and crystallographic compatibility with calcite (the solubility product of HAP at 25 °C being Ksp = 1.6·10−117) (34, 39). The theoretical P content of the ideal hydroxyapatite is approximately 18.5 wt% and the Ca/P molar ratio is close to 1.67. These values significantly differ from what we have observed. However, as already mentioned above, an ideal hydroxyapatite in nature rarely precipitates from the aqueous solutions and therefore it is more common to form hydroxyapatite which is non-stoichiometric and substituted by various ions, such as carbonate ions. Also in this research, the reaction between the DAP and the CaCO3-rich matrices does not involve complete transformations and the overwhelming signal of the unreacted CaCO3 affects the measured Ca to P ratio. XRD analysis of samples taken from the treated with DAP surfaces (for 3 and 6 hours of reactant treatment time) did not yield a positive identification of Ca-P phases owing to the very strong signal of the dominating CaCO3 and/or SiO2 particles used as aggregates in the test blocks. The phosphorous content and profile distribution was also assessed through SEM-EDS on polished cross-sectional areas at the core of the sample. Micrographs and EDS elemental spectra and maps were taken at specific areas along the sample width: a) close to treated surface; b) in the center and c) at the bottom margin (furthest from the surface). Phosphorous (P) maps indicated newly formed calcium phosphate phases located in areas originally occupied by the microcrystalline CaCO3 matrix (Table V(a), V(b), Figures 8 (a,b), 9 (a,b,c) and 10 (a, b,c)). Phosphorus content (measured at 1 μm2 of surface area) ranged between 1 and 7 wt% for samples treated with 1 M DAP (Tables VI and VII) and between 1 and 10 wt% for samples treated with 2 M DAP. Traces of chlorine (Cl) were attributed to the epichlorohydrin in the hardener of the epoxy resin used for the preparation of the cross-sections. The P concentration was found to be higher closer to the surface of the sample which was in contact area with the DAP saturated compresses and in areas surrounded by large pores as more of the DAP solution could enter and fill up nearby pores and react with surrounding CaCO3. As SiO2 particles are not expected to react with the DAP solution, the formation of calcium phosphate phases for the quartz-based test blocks is considered as the result of the DAP interaction with the microcrystalline CaCO3 binder surrounding the quartz aggregate (Figure 10 (a-d)). Since the marble dust is composed of calcite (CaCO3) it is believed that in the case of the marble dust test blocks, the calcite aggregates also partially react with the DAP solution (Figure 9 (a-c)). The extent of this reaction may be kinetically restrained due to the short contact times (between 3-6 hours) of the proposed consolidation treatments. 428 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 4. Surfaces of untreated wall painting test blocks prepared from lime putty and aggregates at 1:4 volumetric ratio; (a) marble dust aggregates; (b) quartz aggregates.
429 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 5. Surfaces of wall painting test blocks prepared from lime putty and aggregates treated with 1M DAP solution for 6 h; (a) samples prepared from marble dust note the formation of calcium phosphate crystals network, points 1 and 2 indicate the areas analyzed using EDS-Table II(a)(a); (b) samples prepared from quartz showing the formation of a calcium phosphate network over the microcrystalline CaCO3 binder in the region where it completely covers SiO2 grains, points 1 and 2 indicate the areas analyzed using EDS-Table II(b)(b).
430 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 6. Surfaces of wall painting test blocks prepared from lime putty and aggregates treated with 2M DAP solution for 3 h; (a) samples prepared from marble dust, points 1, 2 and 3 indicate the areas analyzed with EDS-Table III(a)(a); (b) samples prepared from quartz aggregates, points 1, 2 and 3 indicate areas analyzed using EDS-Table III(b)(b).
431 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 7. Surface of wall painting test block prepared from lime putty and aggregates treated with 2M DAP solution for 3 h showing the formation of different calcium phosphate phases; (a) samples prepared from marble dust aggregates, points 1,2 and 4 show large Ca-P crystals of ~150-250 μm2, points 1-4 indicate the areas analyzed using EDS-Table IV(a)(a); (b) samples prepared from quartz aggregates, points 1, 2, and 3 show areas analyzed using EDS-Table IV(b)(b).
432 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Table II(a). Weight % of elements measured on the surface of the wall painting test block (prepared using marble dust aggregates and treated with 1M DAP solution for 6 h) in areas 1 and 2 as indicated in Figure 5(a) Measured point
C content
O content
P content
Ca content
1M marble-1
7.21
37.40
3.75
51.64
1M marble-2
6.94
38.38
5.57
49.10
Table II(b). Weight % of elements measured on the surface of the of the wall painting test block (prepared using quartz aggregates and treated with 1M DAP solution for 6 h) in areas 1 and 2 as indicated in Figure 5(b) Measured point
C content
O content
Mg content
Si content
P content
Ca content
1M quartz-1
14.72
48.66
0.00
1.00
3.37
32.25
1M quartz-2
14.69
56.06
0.39
1.12
3.15
24.58
Table III(a). Weight % of elements measured on the surface of the wall painting test block in (prepared using marble dust aggregates and treated with 2M DAP solution for 3 h) areas 1, 2 and 3 as indicated in Figure 6(a) Measured point
C content
O content
P content
Ca content
2M marble-1
11.20
29.29
2.05
57.46
2M marble-2
9.05
31.72
2.54
56.69
2M marble-3
16.62
50.91
2.71
29.76
433 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Table III(b). Weight % of elements measured on the surface of the wall painting test block (prepared using quartz aggregates and treated with 2M DAP solution for 3 h) in areas 1, 2 and 3 as indicated in Figure 6(b) Measured point
C content
O content
Al content
Si content
P content
C content
2M quartz-1
28.61
50.10
0.63
18.81
0.00
1.85
2M quartz-2
15.94
51.51
0.00
3.30
1.48
27.77
2M quartz-3
17.39
44.99
0.00
8.87
4.03
24.72
Table IV(a). Weight % of elements measured on the surface of the sample (prepared using marble dust, treated with 2M DAP for 3 h) in areas 1-4 as indicated in Figure 7(a). Points 1, 2, and 4 show large crystals of a different calcium phosphate phase. Measured point
C content
O content
P content
Ca content
2M marble-1
10.32
39.84
17.37
32.46
2M marble-2
9.28
57.66
11.34
21.72
2M marble-3
13.78
48.75
0.76
36.71
2M marble-4
8.76
51.25
15.03
24.95
Table IV(b). Weight % of elements measured on the surface of the sample (prepared using quartz, treated with 2M DAP solution for 3 h) in areas 1, 2 and 3 as indicated in Figure 7(b). Point 1 shows a large crystal of a different calcium phosphate phase. Measured point
C content
O content
Mg content
Al content
Si content
P content
Ca content
2M quartz-1
14.89
51.28
4.43
0.00
1.63
8.85
18.93
2M quartz-2
19.71
40.59
0.00
0.00
2.93
4.62
32.15
2M quartz-3
21.50
43.67
0.46
1.44
13.86
3.56
15.51
To investigate the formation of calcium phosphate phases on marble dust grains an additional experiment was conducted where loose marble dust particles (diameter 0.2-0.6 mm) were placed in a beaker containing 1M DAP (solid/solution ratio 1:10) and left in the solution for 1 week. Subsequently, the solid was filtered, dried in the air for 3 days and analyzed using SEM. The grains showed a platy morphology with large well defined crystals, typical for the hexagonal hydroxyapatite phase (Figure 11) (45) and characteristic for the carbonate substituted hydroxyapatite (46). 434 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 8. SEM-BSE micrograph of a polished section of the wall painting test block prepared from lime putty and aggregates; (a) marble dust aggregates: untreated, points 1-5 indicate the areas analyzed using EDS-Table V(a)(a); (b) quartz aggregates: untreated, points 1,2 and 3 indicate areas analyzed using EDS-Table V(b)(b).
435 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 9. SEM-EDS images of a polished section of a wall painting test block prepared from lime putty and marble dust aggregates and treated for 6 h with 1M DAP solution; (a) points 1-5 indicate areas analyzed using EDS-Table VI; (b) elemental map indicating the spatial distribution of Ca in a cross-sectional view of a test. This image shows that the Ca is located in areas corresponding to the marble dust particles as well as the microcrystalline CaCO3 binding matrix; (c) elemental map indicating the spatial distribution of P in a cross-sectional view of a test block. This image shows that the P is located in areas corresponding to the microcrystalline CaCO3 binding matrix and at the boundaries of the large Ca-rich particles.
436 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
437
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 10. SEM-BSE micrograph of a polished section of a wall painting test block prepared from lime putty and quartz aggregates and treated for 6 h with 1M DAP solution; (a) points 1-3 indicate areas analyzed using EDS-Table VII; (b) elemental map indicating the spatial distribution of Si in a cross-sectional view of a test block. This image shows that the Si is located in areas corresponding to the quartz particles; (c) elemental map indicating the spatial distribution of Ca in a cross-sectional view of a test block. This image shows that the Ca is located in areas corresponding to the microcrystalline CaCO3 binding matrix; (d) elemental map indicating the spatial distribution of P in a cross-sectional view of a test block. This image shows that the P is located in areas corresponding to the microcrystalline CaCO3 binding matrix.
Table V(a). Weight % of elements measured on the surface of the untreated wall painting test block (prepared using marble dust aggregates) in areas 1-5 as indicated in Figure 8(a) Measured point
C content
O content
Mg content
Cl content
Ca content
Untreated marble-1
25.77
43.89
0.25
0.00
30.09
Untreated marble-2
12.71
43.52
0.58
0.00
43.19
Untreated marble-3
32.45
40.72
0.33
0.00
26.50
Untreated marble-4
13.89
46.27
0.47
0.00
39.38
Untreated marble-5
84.01
8.94
0.00
2.33
4.72
Table V(b). Weight % of elements measured on the surface of the untreated wall painting test block (prepared using lime quartz aggregates), in areas 1 and 2 as indicated in Figure 8(b) Measured point
C content
O content
Si content
S content
Cl content
Ca content
Untreated quartz-1
3.05
56.02
40.93
0.00
0.00
0.00
Untreated quartz-2
41.42
36.41
1.14
0.20
0.57
20.27
Untreated quartz-3
32.13
40.95
1.30
0.00
0.00
25.63
Untreated quartz-4
3.80
55.07
41.13
0.00
0.00
0.00
Untreated quartz-5
80.17
15.81
1.68
0.00
1.62
0.73
438 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Table VI. Weight % of elements measured on the surface of the wall painting test block (prepared using marble dust aggregates and treated with 1M DAP solution for 6 h) in areas 1-5 as indicated in Figure 9(a) Measured point
C content
O content
Mg content
P content
Cl content
Ca content
1M marble-1
12.43
37.21
0.14
6.80
0.00
43.42
1M marble-2
11.37
43.20
0.37
0.00
0.00
45.05
1M marble-3
21.16
41.20
0.25
4.83
0.00
32.57
1M marble-4
18.67
39.99
0.00
5.73
0.00
35.61
1M marble-5
84.09
8.84
0.00
0.00
2.80
4.28
Table VII. Weight % of elements measured on the surface of the wall painting test block (prepared using quartz aggregates and treated with 1M DAP solution for 6 h) in areas 1, 2 and 3 as indicated in Figure 10(a) Ca
Measured point
C content
O content
Si content
P content
1M quartz-1
30.66
37.13
1.27
3.98
26.96
1M quartz-2
3.69
52.09
44.23
0.00
0.00
1M quartz-3
24.03
41.48
1.06
4.40
29.02
content
XRF Data (Portable XRF) XRF analysis was performed on both untreated and treated samples. XRF allowed for semi quantitative measurements of the P at relatively large areas of the sample (spot size ~ 8mm). XRF analysis was carried out on the same crosssectional areas of the core of the samples analyzed with the EDS. As with the EDS analysis, each of the sections was measured in 3 locations: at the surface, in the center and at the bottom (away from the surface) of each slice (Figure 12). For each sample an average (mean) of three XRF readings was calculated. It was observed that the P content decreases with increasing distance from the DAP saturated poultice (Figure 13 (a)) while the Ca concentration remained constant throughout the samples (Figure 13 (b)). The results indicated that at the surface (closer to the DAP poultice) the P concentration was the highest. However, even at the bottom of the section (furthest from the surface), phosphorous could be detected. This is an indication that the phosphorous (induced in the DAP solution) was transported for at least 2 cm distance using this application method. As with the EDS analysis, the Cl detected (~0.3-0.5 wt %) was attributed to one of the components of the embedding resin used. P content of the untreated samples was below the limit of detection. 439 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 11. Detail of newly formed Ca-P phases on marble dust particles after immersion in 1M DAP solution for 1 week; note platy morphology typical for carbonate-substituted hydroxyapatite.
Figure 12. Polished cross-section of the sample analyzed using XRF spectroscopy. Elemental analysis (semi-quantitative) was performed on the surface, in the center and the bottom of the sample; white dots show the locations of the measurements. 440 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Figure 13. Mean values (n=3) of phosphorus (a) and calcium; (b) contents measured in wt%, at the surface, in the center and at the bottom of the wall painting test blocks at different consolidation treatments. Water Sorption Test The water sorption tests between untreated and treated samples show a final reduction in water sorptivity in the range of ~8-11 wt%. The results (Figure 14 (a,b)), show an exponential trend with noticeably reduced initial sorptivity in the DAP treated samples, compared to the untreated ones (time measured up to ~5 min). The delay in the water intake at the beginning of the test for the treated samples suggests potential change (reduction) in porosity, inhibiting water from 441 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
entering easily (34). In particular, the 2 M DAP-treated sample exhibited an initial sorption rate markedly lower (between ~ 20-32 wt%) than the untreated sample, possibly as a result of the higher pore system modification that occurred on the surface. For the 2 M DAP treatments, excessive whitening and formation of a thin ‘crust’ on the surface was also detected. This indicated the formation of a much denser layer of calcium phosphates in comparison to the 1 M treatments. However the sorptivity trends > 300 minutes seemed to be similar regardless to the DAP concentration.
Figure 14. Water sorptivity measured for wall painting test blocks prepared from lime putty and (a) marble dust aggregates; (b) quartz aggregates, which undergone different consolidation treatments. 442 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Scotch Tape (Adhesion) Test
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Scotch tape (adhesion) tests were also used to evaluate the performance of the consolidation treatment (Figure 15). The results indicated that fewer particles were separated from the surface for the samples treated with DAP compared to the untreated samples. The mass change results before and after the consolidation are shown in Table VIII(a), VIII(b).
Figure 15. Photograph showing visually the results from the scotch tape test performed on a wall painting test block made with marble dust aggregates: untreated sample (left), 1M DAP treated for 6 h (center) and 2M DAP treated for 6 h (right). The treated samples clearly show a reduction of in the particles adhered to the tape suggesting an effective consolidation treatment.
Table VIII(a). Mass change before and after the consolidation treatment for wall painting test blocks prepared with marble dust aggregates Treatment conditions
1M DAP 3h
1M DAP 6h
2M DAP 3h
2M DAP 6h
% of mass change before and after the consolidation
84.73%
88.18%
89.66%
98.52%
443 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Table VIII(b). Mass change before and after the consolidation treatment for wall painting test blocks prepared with quartz aggregates Treatment conditions
1M DAP 3h
1M DAP 6h
2M DAP 3h
2M DAP 6h
% of mass change before and after the consolidation
68.45%
69.90%
67.48%
83.01%
The smaller the percentage is, the better the consolidation effect of the surface. The best consolidation effect was observed for samples treated for 6 hours with 2 M diammonium hydrogen phosphate solution (DAP) e.g. for marble dust-based wall painting test blocks, 98% fewer particles detached in comparison with the untreated samples. However, excessive whitening of the surface was observed. Whitening is not desired as it aesthetically affects wall painting’s appearance.
Conclusions In situ phosphate-based consolidation utilizing ammonium phosphate precursors is a promising treatment to re-establish cohesion of powdery Ca-rich matrices such as wall paintings (like frescoes). Preliminary results indicated the formation of a porous hydroxyapatite network at the surface and subsurface of the wall painting test blocks; reduction of water absorption and insignificant color change. Less ‘whitening’ of the surface due to the formation of new phases was observed on samples treated with 1 M DAP solutions in comparison with 2 M DAP solutions. It was found that phosphate solution enters deeply into the wall painting test blocks and penetration depth of the proposed consolidation method is at least up to 2 cm. Due to hydroxyapatite’s wide pH stability an improved resistance against low pH (acid rains) is expected. Consolidation treatment seems effective for wall paintings and calcareous stones but is believed to have also potential for archaeological tooth, bone and fossil reconstruction. Investigations on bone consolidation are in progress in the scope of another conservation project. Future Work Future work will involve testing of different application methods (controlled spraying) and pre-treatment with nanoparticle calcium hydroxide, peptides and chitosan. Interactions of mineral and organic pigments such as cinnabar, green earth, malachite, orpiment, lead white, hematite, madder lake will be evaluated. Mechanical properties such as measurement of dynamic elastic modulus as well as porosity will be assessed for untreated and treated wall painting test blocks. Different precursors including monoammonium dihydrogen phosphate (MAP) and their consolidation applicability are currently being evaluated. Thermodynamic equilibrium-based modeling to predict and optimize formation 444 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
of calcium phosphate phases is in progress. Weathering tests as well as resistance towards acid rains and aggressive salts ingress will be performed in the last stage of the project.
Acknowledgments
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
Authors would like to acknowledge the National Science Foundation (Award # 1139227, Solid State and Materials Chemistry program, Division of Materials Research) for the full financial support of this work.
References 1.
2. 3. 4. 5.
6.
7.
8. 9. 10.
11.
12. 13. 14. 15.
Kickelbick, G. In Hybrid Materials. Synthesis, Characterization, and Applications; Wiley-VCH, Verlag GmbH & Co. KGaA: Weinheim, Germany, 2007; pp 1−46. Borgia, G. C.; Camaiti, M.; Cerri, F.; Fantazzini, P.; Piacenti, F. Stud. Conserv. 2003, 48, 217–226. Baglioni, P.; Giorgi, R. Soft Matter 2006, 2, 293–303. Miliani, C.; Velo-Simpson, M. L.; Scherrer, G. W J. Cult. Herit. 2007, 8, 1–6. Dardes, K., Rothe, A., Eds.; The Structural Conservation of Panel Paintings; Proceedings of a symposium at The J. Paul Getty Museum, Los Angeles, April 24−28, 1993. Lanterna, G.; Mairani, A.; Matteini, M.; Rizzi, M.; Scuto, S.; Vincenzi, F.; Zannini, P. In Proceedings of the 9th International Congress on Deterioration and Conservation of Stone; Fassina, V., Ed.; Elsevier: New York, 2000; pp 387−394. Hansen, E.; Doehne, E.; Fidler, J.; Larson, J.; Matteini, M.; RidriguezNavarro, C.; de Tagle, A.; Teutonico, J. M.; Weiss, N. Rev. Conserv. 2003, 4, 13–25. Laurenzi Tabasso, M. APT Bull. 1995, 26, 17–21. Warren, J. In Conservation of Earth Structures; Butterworth-Heinemann: Oxford, U.K., 1999; p 121. Chiari, G. In 6th International Conference on the Conservation of Earthen Architecture; Adobe 90 preprints; Las Cruces, New Mexico, October 14−19, 1990; Grimstad; K., Ed.; The Getty Conservation Institute: Los Angeles, 1990; Vol. 273, pp 267−273. Agnew, N.; Preusser, F.; Druzik, J. R. In 5th International Meeting of Experts on the Conservation of Earthen Architecture; ICCROM: Rome, 1987; pp 75−79. Brus, J.; Kotlik, P. Stud. Conserv. 1996, 41, 109–119. Karatasios, I.; Theoulakis, P.; Kalagri, A.; Sapalidis, A.; Kilikoglou, V. Constr. Build. Mater. 2009, 23, 2803–2812. Wheeler, G. Research in Conservation; The Getty Conservation Institute: Los Angeles, 2005. Carretti, E.; Dei, L.; Baglioni, P. Langmuir 2003, 19, 7867–7872. 445 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
16. Wayne Smith, C. In Archaeological Conservation Using Polymers: Practical Applications for Organic Artifact Stabilization; Texas A&M University Press: College Station, TX, 2003; pp 1−144. 17. Scherrer, G. W.; Wheeler, G. S. Key Eng. Mater. 2009, 391, 1–25. 18. Giorgi, R.; Baglioni, M.; Berti, D.; Baglioni, P Acc. Chem. Res. 2010, 43, 695–704. 19. Ambrosi, M.; Baglioni, P.; Dei, L.; Giorgi, R.; Neto, C. Langmuir 2001, 17, 4251–4255. 20. Baglioni, P.; Giorgi, R. Soft Matter 2006, 2, 293–303. 21. Baglioni, P.; Giorgi, R.; Dei, L. C. R. Chim. 2009, 12, 61–69. 22. Elert, K.; Sebastian, E.; Valverde, I.; Rodriguez-Navarro, C. Appl. Clay Sci. 2008, 39, 122–132. 23. Giorgi, R.; Dei, L.; Baglioni, P. Stud. Conserv. 2000, 45, 154–161. 24. Matteini, M.; Lanterna, G.; Nepoti, M. R. In The Oxalate Films in the Conservation of Works of Art; Proceedings in the II International Symposium; Realini, M., Toniolo, L., Eds.; Milan, Italy, March 25−27, 1996. 25. Matteini, M.; Moles, A.; Giovannoni, S. In The Conservation of Monuments in the Mediterranean Basin. Stone and Monuments: Methodologies for the Analysis of Weathering and Conservation; Proceedings of the 3rd International Symposium; Fassina, V., Ott, H., Zezza, F., Eds.; 1994; pp 155−161. 26. Church, A. H. British Patent 220, 1862. 27. Warren, J. In Conservation of Earth Structures; Butterworth-Heinemann: Oxford, U.K., 1999. 28. Chelazzi, D.; Poggi, G.; Jaidar, Y.; Toccafondi, N.; Giorgi, R.; Baglioni, P. J. Colloid Interface Sci. 2013, 392, 42–49. 29. Hansen, E.; Doehne, E.; Fidler, J.; Larson, J.; Martin, B.; Matteini, M.; Rodriguez-Navarro, C.; Pardo, E. S.; Price, C.; Tagle, A.; Teutonico, J. M.; Weiss, N. Rev. Conserv. 2003, 1–13. 30. Matteini, M. In Conserving the Painted Past: Developing Approaches to Wall Painting Conservation; Post-prints of a conference organized by English Heritage; Gowing, R., Heritage, A., Eds.; James & James: London, 2003; pp 110−115. 31. Matteini, M. In Works of Art & Conservation Science Today; Thessaloniki: Greece, November 26−28, 2010. 32. Yang, F.; Zhang, B.; Liu, Y.; Wei, G.; Zhang, H.; Chen, W.; Xu, Z. New J. Chem. 2011, 35, 887–892. 33. Liu, Q.; Zhang, B. J. Mater. Sci. Forum 2011, 675-677, 317–320. 34. Sassoni, E.; Naidu, S.; Scherer, G. W. J. Cult. Herit. 2011, 12, 346–355. 35. Kamiya, M.; Hatta, J.; Shimida, E.; Ikuma, Y.; Yoshimura, M.; Monma, H. Mater. Sci. Eng. 2004, B 111, 226–231. 36. Matteini, M.; Rescic, S.; Fratini, F.; Botticelli, G. Int. J. Archit. Herit. 2011, 5, 717–736. 37. Steiger, M. Restor. Build. Monum. 2005, 11, 419–431. 38. Scherer, G. W.; Flatt, R; Wheeler, G. MRS Bull. 2001, 26, 44–50. 39. Dorozhkin, S. V. Mater. Struct. 2009, 2, 399–498. 446 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.
Downloaded by ARIZONA STATE UNIV on October 23, 2013 | http://pubs.acs.org Publication Date (Web): October 15, 2013 | doi: 10.1021/bk-2013-1147.ch022
40. Lee, Y.; Hahm, Y. M.; Matsuya, S.; Nakagawa, M.; Ishikawa, K. J. Mater. Sci. 2007, 42, 7843–7849. 41. Wang, L.; Nancollas, G. H. Chem. Rev. 2008, 108, 4628–4669. 42. Zhao, J.; Liu, W.; Zhang, H. Chem. Cent. J. 2011, 5, DOI: 10.1186/1752153X-5-40. 43. Giere, R.; Stille, P. Geol. Sci. 2004, 451, Special Publication 236. 44. Pan, H.; Darvell, B. W. Arch. Oral. Biol. 2007, 52, 618–24. 45. Kobayashi, T.; Ono, S.; Hirakura, S.; Oaki, Y.; Imai, H. Cryst. Eng. Comm. 2012, 14, 1143–1149. 46. Xia, W.; Lin, K.; Gou, Z.; Engqvist, H. In Hydroxyapatite: Synthesis, Properties and Applications; Gshalaev, V. S., Demirchan, A. C., Eds.; Nova Science Publishers, Inc.: New York, 2012; pp 243−264.
447 In Archaeological Chemistry VIII; Armitage, R., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.